Repairs, and what must be disclosed
Updated
Almost every large specimen has been worked on. The question is not whether, but whether it was declared.
Repair is normal
Crystals break in the ground, in the blast, in the bucket and on the bench. Reassembling a specimen that came out in two pieces is ordinary preparation, not deception.
So is removing a mud coating, dissolving a carbonate crust or trimming excess matrix.
The craft has no objection to any of it. It has a strong objection to silence about it.
The spectrum
Cleaning. Removing what was never part of the specimen. Universally accepted, rarely disclosed, and worth noting anyway if it was aggressive.
Trimming. Reducing the matrix. Accepted, normally visible, worth stating.
Repair. Rejoining broken pieces of the same specimen. Accepted with disclosure.
Restoration. Filling a gap with a foreign material, reconstructing a missing part. Accepted only with clear disclosure, and it moves the price materially.
Composite. Assembling a specimen from parts that did not grow together. Disclosed, or it is a fake.
The line between the last two and the first three is the line between preparing and fabricating.
The disclosure rule
Anything done after the specimen came out of the ground that a buyer would want to know: state it on the label and in the description.
The test is what a knowledgeable buyer would consider material, and the safe assumption is that they consider everything material.
Undisclosed repair found later is the fastest way to lose a reputation in a small market where the same few hundred people buy from each other for decades.
Doing it reversibly
Use adhesives that can be removed, and record which one you used on the catalog entry.
A future conservator, or a future owner, needs to be able to take the repair apart without destroying the specimen. An epoxy chosen for strength forecloses that.
And keep the pieces if a repair fails. A specimen in three known pieces is worth more than a specimen in two pieces and a gap.